CN112566117B - Vehicle node identity recognition method and device based on metric learning - Google Patents
Vehicle node identity recognition method and device based on metric learning Download PDFInfo
- Publication number
- CN112566117B CN112566117B CN202011228860.2A CN202011228860A CN112566117B CN 112566117 B CN112566117 B CN 112566117B CN 202011228860 A CN202011228860 A CN 202011228860A CN 112566117 B CN112566117 B CN 112566117B
- Authority
- CN
- China
- Prior art keywords
- wireless signals
- vehicle node
- signals corresponding
- characteristic value
- identity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000004364 calculation method Methods 0.000 claims abstract description 51
- 238000012512 characterization method Methods 0.000 claims abstract description 45
- 238000012360 testing method Methods 0.000 claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 30
- 238000007781 pre-processing Methods 0.000 claims abstract description 24
- 230000006870 function Effects 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 230000009467 reduction Effects 0.000 claims description 9
- 230000001052 transient effect Effects 0.000 claims description 7
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 239000002699 waste material Substances 0.000 abstract description 11
- 238000005259 measurement Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 8
- 238000009826 distribution Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000004075 alteration Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/06—Authentication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/16—Implementing security features at a particular protocol layer
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computer Security & Cryptography (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a vehicle node identity recognition method and device based on metric learning, wherein the method comprises the following steps: establishing a multi-target open type individual characterization calculation model based on a metric learning theory; acquiring wireless signals corresponding to different vehicle nodes and preprocessing the wireless signals to acquire wireless signals corresponding to different vehicle nodes with data dimension N; training a multi-target open individual characterization calculation model according to the wireless signals to obtain a standard characteristic value with the dimension d of the reduced data; inputting wireless signals corresponding to the vehicle nodes to be identified into a trained multi-target open type individual characterization calculation model to obtain a test characteristic value with a data dimension d; calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not; thereby avoiding resource waste while ensuring safety authentication.
Description
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a method for identifying a vehicle node identity for metric learning, a computer readable storage medium, a computer device, and a device for identifying a vehicle node identity for metric learning.
Background
In the related technology, the space-air-ground integrated Internet of vehicles provides effective data resource sharing and fusion by utilizing a space network, improves the cooperative capability of the future vehicle-road-network, and provides important network infrastructure support for comprehensively realizing unmanned vehicle landing application; although the spatial information network plays a significant role in various important application scenes, compared with the traditional network, the space-air-ground integrated network has more serious challenges due to the characteristics of complexity, openness, cross-domain property and the like; therefore, the air-space-ground integrated Internet of vehicles safety protection research has become the key of future intelligent driving popularization, and has attracted wide attention of research scholars at home and abroad.
At present, aiming at the security, more protection technologies focused on a link layer of a network and above, such as a communication standard protocol, a cryptographic encryption algorithm and the like are adopted; however, in the air-ground integrated internet of vehicles, due to the complexity of nodes and the convergence of multiple domains, extra calculation and communication overhead are required to be added in a link layer and above protocols, so that a great amount of resource waste is caused.
Disclosure of Invention
The present invention aims to solve at least to some extent one of the technical problems in the above-described technology. Therefore, an object of the present invention is to provide a vehicle node identity recognition method based on metric learning, which uses the metric learning theory to perform physical layer security authentication to realize multi-objective open class recognition, so that extra calculation and communication overhead are not required to be added in a link layer or above protocol, thereby avoiding resource waste while ensuring security authentication.
A second object of the present invention is to propose a computer readable storage medium.
A third object of the invention is to propose a computer device.
A fourth object of the present invention is to provide a vehicle node identification device based on metric learning. In order to achieve the above object, an embodiment of a first aspect of the present invention provides a vehicle node identification method based on metric learning, including the following steps: establishing a multi-target open type individual characterization calculation model SigTLNet based on a metric learning theory; acquiring wireless signals corresponding to different vehicle nodes, and taking the wireless signals as identity characteristics of the corresponding vehicle nodes; preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension of N; training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with the data dimension of N so as to obtain a standard characteristic value with the data dimension of d after dimension reduction; acquiring a wireless signal corresponding to a vehicle node to be identified, and inputting the wireless signal corresponding to the vehicle node to be identified into a trained multi-target open type individual characterization calculation model SigTLNet so as to obtain a test characteristic value with a data dimension of d; and calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not.
According to the vehicle node identity recognition method based on metric learning, firstly, a multi-target open type individual characterization calculation model SigTLNet is established based on a metric learning theory; acquiring wireless signals corresponding to different vehicle nodes, taking the wireless signals as identity characteristics of the corresponding vehicle nodes, preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension N, training a multi-target open type individual characterization calculation model SigTLNet according to the wireless signals corresponding to the different vehicle nodes with the data dimension N to obtain a standard characteristic value with the data dimension d after dimension reduction, acquiring the wireless signals corresponding to the vehicle nodes to be identified, and inputting the wireless signals corresponding to the vehicle nodes to be identified into a trained multi-target open type individual characterization calculation model SigTLNet to obtain a test characteristic value with the data dimension d; finally, calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not; therefore, the physical layer safety authentication is carried out by utilizing the measurement learning theory, and the multi-target open class identification is realized, so that extra calculation and communication expenditure are not required to be added in a link layer or above protocol, and the resource waste is avoided while the safety authentication is ensured.
In addition, the vehicle node identification method based on metric learning according to the above embodiment of the present invention may further have the following additional technical features:
optionally, when a multi-target open individual characterization calculation model SigTLNet is established, an identity recognition module is constructed by adopting a multi-layer deep convolutional neural network on the basis of an acceptance network model structure, wherein a loss function adopts improved triplet loss, a training process adopts an Adam optimizer, and gradient clipping is introduced in training.
Optionally, preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension of N, including: determining the starting point of the wireless signal by adopting a phase method, and extracting envelope information of the wireless signal to obtain the wireless signal with the envelope information reserved for the sampling point, wherein the wireless signal is a transient signal; and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
Optionally, comparing the minimum euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test feature value is legal, including: and if the minimum Euclidean distance is smaller than the preset threshold value, the identity of the vehicle node to be identified corresponding to the test feature value is identified to be legal.
To achieve the above object, a second aspect of the present invention provides a computer-readable storage medium having stored thereon a vehicle node identification program based on metric learning, which when executed by a processor implements a vehicle node identification method based on metric learning as described above.
According to the computer readable storage medium, the vehicle node identification program based on the measurement learning is stored, so that the processor realizes the vehicle node identification method based on the measurement learning when executing the vehicle node identification program based on the measurement learning, and the resource waste is avoided while the safety authentication is ensured.
To achieve the above object, an embodiment of a third aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the method for identifying a vehicle node based on metric learning as described above when the processor executes the program.
According to the computer equipment provided by the embodiment of the invention, firstly, the multi-target open type individual characterization calculation model SigTLNet is established based on the metric learning theory, and the vehicle node identity recognition model program based on the metric learning is stored through the memory, so that the vehicle node identity recognition method based on the metric learning is realized when the vehicle node identity recognition model program based on the metric learning is executed by the processor, and the safety authentication is ensured and the resource waste is avoided.
To achieve the above object, a fourth aspect of the present invention provides a vehicle node identification device based on metric learning, including: the model building module is used for building a multi-target open type individual characterization calculation model SigTLNet, and training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with the data dimension of N so as to obtain a standard characteristic value with the data dimension of d after dimension reduction; the acquisition module is used for acquiring wireless signals corresponding to different vehicle nodes and taking the wireless signals as identity characteristics of the corresponding vehicle nodes; the preprocessing module is used for preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension of N; the training module is used for acquiring wireless signals corresponding to the vehicle nodes to be identified, and inputting the wireless signals corresponding to the vehicle nodes to be identified into the trained multi-target open type individual characterization calculation model SigTLNet so as to obtain test characteristics with the data dimension d; the detection and identification module is used for calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not.
According to the vehicle node identity recognition device based on the metric learning, a model building module is used for building a multi-target open type individual characterization calculation model SigTLNet based on the metric learning theory, and training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with data dimension N to obtain a standard characteristic value with data dimension d after dimension reduction; then, acquiring wireless signals corresponding to different vehicle nodes through an acquisition module, and taking the wireless signals as identity characteristics of the corresponding vehicle nodes; then, preprocessing the wireless signals corresponding to different vehicle nodes through a preprocessing module to obtain wireless signals corresponding to different vehicle nodes with data dimension N; the wireless signals corresponding to the vehicle nodes to be identified are obtained through a training module, and are input into a trained multi-target open type individual characterization calculation model SigTLNet, so that test characteristics with the data dimension d are obtained; finally, calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value through the detection and identification module, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not; therefore, the physical layer safety authentication is carried out by utilizing the measurement learning theory, and the multi-target open class identification is realized, so that extra calculation and communication expenditure are not required to be added in a link layer or above protocol, and the resource waste is avoided while the safety authentication is ensured.
Optionally, the preprocessing module is further configured to: determining the starting point of the wireless signal by adopting a phase method, and extracting envelope information of the wireless signal to obtain the wireless signal with the envelope information reserved for the sampling point, wherein the wireless signal is a transient signal; and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
Optionally, when a multi-target open individual characterization calculation model SigTLNet is established, an identity recognition module is constructed by adopting a multi-layer deep convolutional neural network on the basis of an acceptance network model structure, wherein a loss function adopts improved triplet loss, a training process adopts an Adam optimizer, and gradient clipping is introduced in training.
Optionally, the detection and identification module is further configured to: and if the minimum Euclidean distance is smaller than the preset threshold value, the identity of the vehicle node to be identified corresponding to the test feature value is identified to be legal.
Drawings
FIG. 1 is a flow chart of a vehicle node identification method based on metric learning according to an embodiment of the invention;
FIG. 2 is a system for extracting and identifying physical fingerprint features of a wireless signal;
FIG. 3 is a diagram of different types of sample signals after preprocessing;
FIG. 4 is a schematic diagram of an identification model structure;
FIG. 5 is a comparison of the original triplet loss versus the improved triplet loss convergence;
FIG. 6 is a comparison of sample feature distribution extracted by an identity recognition model with sample feature distribution extracted by PCA;
FIG. 7 is a graph showing the change of network loss values after increasing gradient clipping;
FIG. 8 is a graph showing the variation of the recognition rate of the identification model with the threshold value;
FIG. 9 is a graph showing the comparison of recognition rates under different signal-to-noise ratios;
fig. 10 is a block diagram of a vehicle node identification device based on metric learning according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present invention and should not be construed as limiting the invention.
In order that the above-described aspects may be better understood, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Fig. 1 is a flow chart of a vehicle node identity recognition method based on metric learning according to an embodiment of the present invention, as shown in fig. 1, the vehicle node identity recognition method based on metric learning according to an embodiment of the present invention includes the following steps:
step 101, a multi-target open class individual characterization calculation model SigTLNet is established based on a metric learning theory.
That is, a multi-objective open class individual characterization calculation model SigTLNet needs to be constructed first.
As an embodiment, as shown in fig. 4, when a multi-target open type individual characterization calculation model SigTLNet is established, an identity recognition module is constructed by adopting a multi-layer deep convolution neural network on the basis of an acceptance network model structure, wherein a loss function adopts improved triplet loss, a training process adopts an Adam optimizer, and gradient clipping is introduced in training.
As a specific embodiment, the loss function adopts improved triplet loss, and the formula is as follows: thereby accelerating the convergence speed; wherein M represents the number of triplets, i=1, 2 … M, a, p, n are triplets, a and p are the same type of signals, and a and n are different types of signals; alpha is a preset parameter, in this embodiment 0.2; + indicates that when the value in the bracket is greater than zero, the value is taken as a loss, and when the value in the bracket is less than zero, the loss is zero.
Fig. 5 shows the original triplet loss versus the improved triplet loss convergence, and it can be seen that the improved triplet loss can significantly accelerate the convergence.
Fig. 6 is a comparison of sample feature distribution extracted by the multi-target open class individual characterization calculation model SigTLNet and sample feature distribution extracted by the PCA, and compared with the PCA, the feature distribution among the sample classes extracted by the multi-target open class individual characterization calculation model SigTLNet is more separated.
Fig. 7 shows a change of a network loss value after gradient clipping, gradient explosion can be developed (the loss value becomes NAN) in the training process without gradient clipping, and the loss function can be converged normally after gradient clipping is introduced.
Step 102, acquiring wireless signals corresponding to different vehicle nodes, and taking the wireless signals as the identity characteristics of the corresponding vehicle nodes.
As an embodiment, the wireless signal collection of different vehicle nodes is performed by the collection system, for example, the signal collection display is performed by an oscilloscope, which is not particularly limited in the present invention.
And 103, preprocessing the wireless signals corresponding to the different vehicle nodes to obtain the wireless signals corresponding to the different vehicle nodes with the data dimension of N.
As an embodiment, the preprocessing process includes origin detection and normalization of the signal, for example: determining a starting point of a wireless signal by adopting a phase method, and extracting envelope information of the wireless signal to obtain the wireless signal with the envelope information reserved at a sampling point corresponding to the envelope information, wherein the wireless signal is a transient signal; and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
That is, a phase method is adopted to determine the starting point of the transient signal, envelope information is extracted, and 14400 sampling points are reserved on the final signal; and further denoising and normalizing the signals.
It should be noted that, in order to verify the feasibility of the network model based on measurement deep learning proposed herein, the verification is performed through wireless signal data of different types of nodes such as actually collected unmanned aerial vehicles, mobile phones, ioT devices, ships and the like; meanwhile, a simulation experiment is carried out in combination with the identification and authentication process of the legitimacy of the identity information of the vehicle node in the space-earth integrated network, and the simulation experiment is compared with the traditional cluster identification algorithm, so that the multi-target open type individual identification is realized, and the identification accuracy under different signal-to-noise ratio conditions is verified; wherein the different types of sample signals after preprocessing are shown in fig. 3.
And 104, training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with the data dimension of N so as to obtain a standard characteristic value with the data dimension of d after dimension reduction.
That is, the preprocessed wireless signals x= [ X ] corresponding to different vehicle nodes with data dimension N 1 ,x 2 ,…x N ]Inputting the D-dimension standard characteristic value into a constructed identity recognition module for training to obtain a D-dimension standard characteristic value after the training
Step 105, acquiring a wireless signal corresponding to the vehicle node to be identified, and inputting the wireless signal corresponding to the vehicle node to be identified into the trained identity recognition model to obtain a test characteristic value with the data dimension d.
That is, the wireless signal X' of the test vehicle node is collected and sent into a trained multi-target open type individual characterization calculation model SigTLNet to obtain a characteristic value f D ′。
And 106, calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not.
As one example, f 'is calculated by the following formula' d Minimum Euclidean distance from standard eigenvalue F Wherein f i ' represents the ith test feature value, f i j Representing the mth standard characteristic value.
As an embodiment, as shown in fig. 2, if the minimum euclidean distance dis is greater than a preset threshold, the identity of the vehicle node to be identified corresponding to the test feature value is not legal, and if the minimum euclidean distance dis is less than the preset threshold, the identity of the vehicle node to be identified corresponding to the test feature value is legal.
Fig. 8 shows the change condition of the network recognition rate of the multi-target open individual characterization calculation model SigTLNet along with a preset threshold, wherein the preset threshold is set to be 0.8.
FIG. 9 is a graph showing the comparison of recognition rates under different signal-to-noise ratios; under the condition of SNR=30 dB, the recognition rate of both the SigTLNet and the KNN can reach more than 94%, but the SigTLNet can be used for unknown targets, and the KNN can only be used for closed set recognition; the K-Means clustering has poor wireless signal classification recognition results with small differences; under the condition of poor signal-to-noise ratio, namely snr=5db, the overall recognition accuracy of the sigtlnet network is higher compared with the KNN and K-Means algorithm.
In summary, the invention provides a vehicle node identity recognition method based on metric learning, which utilizes an improved triplet metric loss function to establish a multi-target open class individual characterization calculation model SigTLNet; by judging the signal characteristic distance, legal and malicious node identities in the multi-target open class are effectively identified, the problems that the traditional identification method is wasteful in extracting and establishing a sample database storage space, difficult in acquiring complete closed set data and the like are solved, and meanwhile, manual characteristic fingerprint selection is not needed according to different node wireless signals.
According to the vehicle node identity recognition method based on metric learning, firstly, a multi-target open type individual characterization calculation model SigTLNet is established based on a metric learning theory; acquiring wireless signals corresponding to different vehicle nodes, taking the wireless signals as identity characteristics of the corresponding vehicle nodes, preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension N, training a multi-target open type individual characterization calculation model SigTLNet according to the wireless signals corresponding to the different vehicle nodes with the data dimension N to obtain a standard characteristic value with the data dimension d after dimension reduction, acquiring the wireless signals corresponding to the vehicle nodes to be identified, and inputting the wireless signals corresponding to the vehicle nodes to be identified into a trained multi-target open type individual characterization calculation model SigTLNet to obtain a test characteristic value with the data dimension d; finally, calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not; therefore, the physical layer safety authentication is carried out by utilizing the measurement learning theory, and the multi-target open class identification is realized, so that extra calculation and communication expenditure are not required to be added in a link layer or above protocol, and the resource waste is avoided while the safety authentication is ensured.
In addition, the embodiment of the invention also provides a computer readable storage medium, on which a vehicle node identification program based on metric learning is stored, and the vehicle node identification program based on metric learning realizes the vehicle node identification method based on metric learning when being executed by a processor.
According to the computer readable storage medium, the vehicle node identification program based on the measurement learning is stored, so that the processor realizes the vehicle node identification method based on the measurement learning when executing the vehicle node identification program based on the measurement learning, and the resource waste is avoided while the safety authentication is ensured.
In addition, the embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the vehicle node identity recognition method based on the measurement learning when executing the program.
According to the computer equipment provided by the embodiment of the invention, the vehicle node identity recognition program based on measurement learning is stored in the memory, so that multi-target open class recognition can be realized; therefore, the vehicle node identity recognition method based on the measurement learning is realized when the vehicle node identity recognition program based on the measurement learning is executed by the processor, so that the safety authentication is ensured and the resource waste is avoided.
FIG. 10 is a block diagram of a vehicle node identification apparatus based on metric learning according to an embodiment of the present invention; as shown in fig. 10, the apparatus includes: the system comprises a model building module 201, an acquisition module 202, a preprocessing module 203, a training module 204 and a detection and identification module 205.
The model building module 201 is configured to build a multi-target open type individual characterization calculation model SigTLNet, and train the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with a data dimension of N, so as to obtain a standard feature value with a data dimension of d after dimension reduction; the acquiring module 202 is configured to acquire wireless signals corresponding to different vehicle nodes, and take the wireless signals as identity features of the corresponding vehicle nodes; the preprocessing module 203 is configured to preprocess wireless signals corresponding to different vehicle nodes to obtain wireless signals corresponding to different vehicle nodes with data dimension N; the training module 204 is configured to obtain a wireless signal corresponding to a vehicle node to be identified, and input the wireless signal corresponding to the vehicle node to be identified into a trained multi-target open class individual characterization calculation model SigTLNet to obtain a test feature with a data dimension d; the detection and identification module 205 is configured to calculate a minimum euclidean distance between the test feature value and the standard feature value, and compare the minimum euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test feature value is legal.
Further, the preprocessing module 203 is further configured to determine a start point of a wireless signal by using a phase method, and extract envelope information of the wireless signal to obtain a wireless signal with a sampling point corresponding to the reserved envelope information, where the wireless signal is a transient signal; and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
Further, when a multi-target open type individual characterization calculation model SigTLNet is established, an identity recognition module is built by adopting a multi-layer deep convolution neural network on the basis of an acceptance network model structure, wherein a loss function adopts triplet loss, a training process adopts an Adam optimizer, and gradient clipping is introduced in training.
Further, the detection and identification module 205 is further configured to identify that the identity of the vehicle node to be identified corresponding to the test feature value is illegal if the minimum euclidean distance is greater than a preset threshold value, and identify that the identity of the vehicle node to be identified corresponding to the test feature value is illegal if the minimum euclidean distance is less than the preset threshold value.
It should be noted that the foregoing description of the vehicle node identification method based on metric learning in fig. 1 is also applicable to the vehicle node identification device based on metric learning, and will not be repeated herein.
In summary, according to the vehicle node identity recognition device based on metric learning in the embodiment of the present invention, a model building module builds a multi-target open type individual characterization calculation model SigTLNet based on a metric learning theory, and trains the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with data dimension N, so as to obtain a standard feature value with dimension d of the reduced data dimension; then, acquiring wireless signals corresponding to different vehicle nodes through an acquisition module, and taking the wireless signals as identity characteristics of the corresponding vehicle nodes; then, preprocessing the wireless signals corresponding to different vehicle nodes through a preprocessing module to obtain wireless signals corresponding to different vehicle nodes with data dimension N; the wireless signals corresponding to the vehicle nodes to be identified are obtained through a training module, and are input into a trained multi-target open type individual characterization calculation model SigTLNet, so that test characteristics with the data dimension d are obtained; finally, calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value through the detection and identification module, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not; therefore, the physical layer safety authentication is carried out by utilizing the measurement learning theory, and the multi-target open class identification is realized, so that extra calculation and communication expenditure are not required to be added in a link layer or above protocol, and the resource waste is avoided while the safety authentication is ensured.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third, etc. do not denote any order. These words may be interpreted as names.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
In the description of the present invention, it should be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
In the present invention, unless explicitly specified and limited otherwise, the terms "mounted," "connected," "secured," and the like are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally formed; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communicated with the inside of two elements or the interaction relationship of the two elements. The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art according to the specific circumstances.
In the present invention, unless expressly stated or limited otherwise, a first feature "up" or "down" a second feature may be the first and second features in direct contact, or the first and second features in indirect contact via an intervening medium. Moreover, a first feature being "above," "over" and "on" a second feature may be a first feature being directly above or obliquely above the second feature, or simply indicating that the first feature is level higher than the second feature. The first feature being "under", "below" and "beneath" the second feature may be the first feature being directly under or obliquely below the second feature, or simply indicating that the first feature is less level than the second feature.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms should not be understood as necessarily being directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
While embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the invention, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the invention.
Claims (8)
1. The vehicle node identification method based on metric learning is characterized by comprising the following steps of:
establishing a multi-target open type individual characterization calculation model SigTLNet based on a metric learning theory, and constructing an identity recognition module by adopting a multi-layer deep convolution neural network on the basis of an acceptance network model structure when the multi-target open type individual characterization calculation model SigTLNet is established, wherein a loss function adopts improved triplet loss, a training process adopts an Adam optimizer, and gradient cutting is introduced in training;
acquiring wireless signals corresponding to different vehicle nodes, and taking the wireless signals as identity characteristics of the corresponding vehicle nodes;
preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension of N;
training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with the data dimension of N so as to obtain a standard characteristic value with the data dimension of d after dimension reduction;
acquiring a wireless signal corresponding to a vehicle node to be identified, and inputting the wireless signal corresponding to the vehicle node to be identified into a trained multi-target open type individual characterization calculation model SigTLNet so as to obtain a test characteristic value with a data dimension of d;
and calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not.
2. The method for identifying the identity of the vehicle node based on metric learning of claim 1, wherein preprocessing the wireless signals corresponding to the different vehicle nodes to obtain the wireless signals corresponding to the different vehicle nodes with the data dimension of N comprises:
determining the starting point of the wireless signal by adopting a phase method, and extracting envelope information of the wireless signal to obtain the wireless signal with the envelope information reserved for the sampling point, wherein the wireless signal is a transient signal;
and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
3. The vehicle node identity recognition method based on metric learning of claim 1, wherein comparing the minimum euclidean distance with a preset threshold value to recognize whether the identity of the vehicle node to be recognized corresponding to the test feature value is legal or not, comprises:
and if the minimum Euclidean distance is smaller than the preset threshold value, the identity of the vehicle node to be identified corresponding to the test feature value is identified to be legal.
4. A computer readable storage medium, characterized in that a vehicle node identification program based on metric learning is stored thereon, which, when executed by a processor, implements the vehicle node identification method based on metric learning as claimed in any one of claims 1-3.
5. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the metric learning based vehicle node identification method of any of claims 1-3 when the program is executed by the processor.
6. A vehicle node identification device based on metric learning, comprising:
the model building module is used for building a multi-target open type individual characterization calculation model SigTLNet, when the multi-target open type individual characterization calculation model SigTLNet is built, an identity recognition module is built by adopting a multi-layer deep convolution neural network on the basis of an acceptance network model structure, wherein a loss function adopts improved triplet loss, a training process adopts an Adam optimizer, and gradient cutting is introduced into training; training the multi-target open type individual characterization calculation model SigTLNet according to wireless signals corresponding to different vehicle nodes with the data dimension of N to obtain a standard characteristic value with the data dimension of d after dimension reduction;
the acquisition module is used for acquiring wireless signals corresponding to different vehicle nodes and taking the wireless signals as identity characteristics of the corresponding vehicle nodes;
the preprocessing module is used for preprocessing the wireless signals corresponding to the different vehicle nodes to obtain wireless signals corresponding to the different vehicle nodes with the data dimension of N;
the training module is used for acquiring wireless signals corresponding to the vehicle nodes to be identified, and inputting the wireless signals corresponding to the vehicle nodes to be identified into the trained multi-target open type individual characterization calculation model SigTLNet so as to obtain test characteristics with the data dimension d;
the detection and identification module is used for calculating the minimum Euclidean distance between the test characteristic value and the standard characteristic value, and comparing the minimum Euclidean distance with a preset threshold value to identify whether the identity of the vehicle node to be identified corresponding to the test characteristic value is legal or not.
7. The metric learning based vehicle node identification device of claim 6, wherein the preprocessing module is further configured to:
determining the starting point of the wireless signal by adopting a phase method, and extracting envelope information of the wireless signal to obtain the wireless signal with the envelope information reserved for the sampling point, wherein the wireless signal is a transient signal;
and denoising and normalizing the wireless signals with the reserved envelope information corresponding to the sampling points to obtain the wireless signals corresponding to different vehicle nodes with the data dimension of N after processing.
8. The metric learning based vehicle node identification device of claim 6, wherein the detection and identification module is further configured to:
and if the minimum Euclidean distance is smaller than the preset threshold value, the identity of the vehicle node to be identified corresponding to the test feature value is identified to be legal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011228860.2A CN112566117B (en) | 2020-11-06 | 2020-11-06 | Vehicle node identity recognition method and device based on metric learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011228860.2A CN112566117B (en) | 2020-11-06 | 2020-11-06 | Vehicle node identity recognition method and device based on metric learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112566117A CN112566117A (en) | 2021-03-26 |
| CN112566117B true CN112566117B (en) | 2023-12-08 |
Family
ID=75041738
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011228860.2A Active CN112566117B (en) | 2020-11-06 | 2020-11-06 | Vehicle node identity recognition method and device based on metric learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112566117B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115514499B (en) * | 2022-11-18 | 2023-03-14 | 广州优刻谷科技有限公司 | Safety communication method, device and storage medium based on mathematical statistics |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013101052A1 (en) * | 2011-12-29 | 2013-07-04 | Intel Corporation | Systems, methods, and apparatus for learning the identity of an occupant of a vehicle |
| CN110399828A (en) * | 2019-07-23 | 2019-11-01 | 吉林大学 | A kind of vehicle recognition methods again of the depth convolutional neural networks based on multi-angle |
| CN110516703A (en) * | 2019-07-18 | 2019-11-29 | 平安科技（深圳）有限公司 | Vehicle identification method, device and storage medium based on artificial intelligence |
| CN111079602A (en) * | 2019-12-06 | 2020-04-28 | 长沙千视通智能科技有限公司 | Vehicle fine granularity identification method and device based on multi-scale regional feature constraint |
| CN111814584A (en) * | 2020-06-18 | 2020-10-23 | 北京交通大学 | Vehicle weight identification method under multi-view-angle environment based on multi-center measurement loss |
-
2020
- 2020-11-06 CN CN202011228860.2A patent/CN112566117B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013101052A1 (en) * | 2011-12-29 | 2013-07-04 | Intel Corporation | Systems, methods, and apparatus for learning the identity of an occupant of a vehicle |
| CN110516703A (en) * | 2019-07-18 | 2019-11-29 | 平安科技（深圳）有限公司 | Vehicle identification method, device and storage medium based on artificial intelligence |
| CN110399828A (en) * | 2019-07-23 | 2019-11-01 | 吉林大学 | A kind of vehicle recognition methods again of the depth convolutional neural networks based on multi-angle |
| CN111079602A (en) * | 2019-12-06 | 2020-04-28 | 长沙千视通智能科技有限公司 | Vehicle fine granularity identification method and device based on multi-scale regional feature constraint |
| CN111814584A (en) * | 2020-06-18 | 2020-10-23 | 北京交通大学 | Vehicle weight identification method under multi-view-angle environment based on multi-center measurement loss |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112566117A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106845339B (en) | Mobile phone individual identification method based on bispectrum and EMD fusion characteristics | |
| CN108390869B (en) | Vehicle-mounted intelligent gateway device integrating deep learning and command sequence detection method thereof | |
| CN110636048B (en) | Vehicle-mounted intrusion detection method and system based on ECU signal characteristic identifier | |
| CN110381509B (en) | Combined authentication method and server suitable for dynamic connection scene | |
| CN111144522A (en) | Power grid NFC equipment fingerprint authentication method based on hardware intrinsic difference | |
| CN112566117B (en) | Vehicle node identity recognition method and device based on metric learning | |
| CN114387627A (en) | Small sample wireless device radio frequency fingerprint identification method and device based on depth measurement learning | |
| CN114900331B (en) | Vehicle-mounted CAN bus intrusion detection method based on CAN message characteristics | |
| CN112491677B (en) | CAN bus identification method and device based on physical layer characteristic fingerprint | |
| Gul et al. | Fine-grained augmentation for RF fingerprinting under impaired channels | |
| CN114867018A (en) | Device identification authentication method, system and storage medium using carrier frequency offset | |
| Kneib et al. | On the robustness of signal characteristic-based sender identification | |
| Aminuddin et al. | Securing wireless communication using RF fingerprinting | |
| CN117560104A (en) | Construction method of interpretable machine learning-assisted channel model in mixed traffic | |
| CN116776227A (en) | Satellite identity recognition method and device based on feature fusion | |
| CN109587136B (en) | Radio frequency fingerprint feature extraction and identification method based on double maximum values | |
| CN115809426A (en) | Radiation source individual identification method and system | |
| CN113420791B (en) | Access control method and device for edge network equipment and terminal equipment | |
| CN115996148A (en) | MIMO OFDM communication-oriented electric power Internet of things physical layer equipment authentication method | |
| CN113365273A (en) | Packet-level wireless equipment authentication method based on channel state information | |
| CN112016539A (en) | Signal identification method and device, electronic equipment and storage medium | |
| CN117896176B (en) | Learning-driven physical layer authentication method for industrial Internet of things spoofing attack | |
| CN114501446B (en) | Physical layer authentication method based on PU (polyurethane) bagging strategy in dynamic industrial scene | |
| CN114997299B (en) | Radio frequency fingerprint identification method in resource limited environment | |
| CN117150265B (en) | Robust radio frequency signal open set identification method under low signal-to-noise ratio condition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |